IT is impossible for anyone who was a pupil of Malinowski to write about his work quite

impersonally. One has to be able to visualize the histrionic, not to say the exhibitionistic, streak
in him to understand the tone of some of his later books. It arose from his view of himself as
the leader of a revolutionary movement in anthropology; and such was his magnetism, his wit
and his virtuosity that he made us, his pupils, fall in eagerly with that view. This happened in
spite of his sometimes offensive prejudices and his impatience of criticism; for he was basically
right. But it warped his work. He could not shake off the compulsion to present his theories
and his ethnographic discoveries in the form of an assault on the ancien regime. It drove him to
wrap up some of his most original ideas and observations in laboured paradoxes and prolix
repetition. Coral Gardens (1935) illustrates this. I mention it in particular because the typescript
was discussed, page by page, in the seminar of 1932-3 of which I was a member. As the Preface
implies, Malinowski regarded it as the Summa Ethnographica of the Trobrianders and, almost
by corollary, as the best example of functionalism in action. Malinowski did indeed strain
himself and his seminar to make it such an example; and that is perhaps why it shows up so
well the blind spots due to his preoccupation with his war against obsolete theories and
imputed opposition. Not surprisingly, this came into almost everything Malinowski wrote on
kinship; for kinship is a subject that bristles with temptations for the anthropologist who feels
strongly about slipshod thinking and well- rooted fallacies. Even today Morgan's preposterous
scheme of the evolution of the family through the stages of promiscuity, group marriage,
matriarchy and patriarchy, is still accepted as a sociological law, and not only in the U.S.S.R.
Though no reputable anthropologist in this country or America now accepts it, non-
anthropological scholars do. One recent example is that of a distinguished classical scholar's
reconstruction of pre-Aegean Greek social organization on the basis of Morgan's and Engels's
theories (Thomson, 1949). Another example is the following statement by one of the most
eminent biologists and sinologists of this century, Joseph Needham, writing about Chinese
social organization in Shang times: 'Matriarchal traces, noted by many authorities, seem to give
place during this period to a rigidly patriarchal society' (1954, P- 85). For Malinowski and his
contemporaries who agreed with him, these temptations were much stronger than for us, their
heirs. They were brought up on the Matriarchal Controversy; and the pseudo-historical and
pseudo-psychological theories of classificatory kinship, incest, totemic descent and so forth
which have now been relegated to the limbo of historical curiosities as a resuU of their
exposure by Malinowski and Radcliffe-Brown, had powerful support as late as the
nineteenthirties. One of the worst offenders was Sir James Frazer. Malinowski's training in
sociology and anthropology was completed under Westermarck's direction and this set the
pattern of his interests in the field of kinship for the rest of his life. It also involved him in the
controversies of the turn of the century when scholars were still vehemently debating whether
the family or the clan came first in man's cultural evolution, and whether 'group marriage'
could be demonstrated from Australian sex customs. Malinowski accepted the theory put
forward with such massive documentation by Westermarck that the individual family is a
universal feature of human life and is also prior to the clan in the historical sense. His first
book, The Family Among the Australian Aborigines (1913a), was a notable, contribution on
Westermarck's side. He often recurred to the absurdity of the 'group marriage' theory as an
instance of how misleading it can be to ignore the total institutional setting of marriage and
sex. But he might well have thrown aside these and related controversies in his Trobriand
studies if it had not been for W. H. R. Rivers and what he stood for in British anthropology. As
the founder of field research on primitive kinship, Rivers still deserves our homage. He was at
the height of his fame and influence during the very years when Malinowski was carrying out
his field-work in the Trobriand islands and was publishing his first results. It is a sign, both of
Rivers's standing, and of the state of British anthropology up to the time of the publication of
The Argonauts of the Western Pacific (1922a), that Rivers was awarded the Gold Medal of the
Royal Society in 191 5, mainly for The History of Melanesian Society (1914). This monument of
bizarre logic and misunderstood ethnology showed, however, that Rivers was incapable of
thinking sociologically about kinship even though he advocated what seemed to be a
sociological theory of the subject. In his early contributions to the Reports of the Torres Straits
Expedition (1904), in his remarkable field report on The Todas (1906), and in his lectures on
Kinship and Social Organization (191 3) which Malinowski attended before he went to the
Pacific, Rivers purported to be seeking for the 'social conditions' that, in his view, 'determine'
'kinship systems', that is, systems of relationship terminology. In fact, the 'social conditions' he
invoked, in the first place, were speculative rules of marriage supposed to have prevailed in the
past; and his shrewd field observations of kinship custom were often twisted to fit his
conjectures. The basic assumption was one that went back to Morgan, who considered kinship
terminologies to be fossil relics of earlier stages of social organization; and not even Radcliffe-
Brown's cogent persuasions (1952, p. 51) moved Rivers from his position. Taking relationship
terminologies as his primary data he sought in them evidence of determination of kinship
relationships by obsolete marriage rules. The much discussed classificatory nomenclature he
was ready to explain as relics of an earlier stage of sexual communism. He had critics but they
differed from him only in what they read into the terminologies. What Rivers stood for
accounts for much of Malinowski's over- emphasis on the 'biographical method' and especially
for the scorn he poured on 'kinship algebra'. He probably felt the more strongly in these
matters just because Rivers was regarded as an authority on Melanesian social organization.
And it went deeper than a disagreement over a point of theory. Kinship terminologies were
usually quoted by him as in- stances of what he contemptuously called linguistic
'coUectioneering'. His pragmatic theory of language with its exaggeration of the instrumental
meaning of words, and his 'conditioning' model of language learning, were- in part a reaction
against the prevailing obsession with kinship terminology.^ Malinowski was much given to
paradox in order to bring out the point of an argument. In his later analysis of the relation
between language and culture (1935, Vol. II) he writes almost as if he sees an antithesis
between 'language' as a mere tissue of words, and 'culture' as the tangible 'reality' of social
life. This is, however, only a special instance of an ingrained mode of thought in Mahnowski's
work. He claimed that functionalism was a '. . . theory which, begun in field-work, leads back to
field-work again' (1932a, p. xxix) and by field-work he meant observation of '. . . what exists . . .
how it works, and what it means to the natives' (ibid., xxxi), in contrast to the hypotheses built
on hearsay and on statements from informants. This general approach was the basis of Crime
and Custom (19266) where it is most forcefully expounded by reference to a problem of
kinship, the Trobriand law of exogamy. What Malinowski tried to drive home—and this is a
recurrent theme in the book—is the discrepancy between 'the ideal of native law' and the
'application of morality and ideals to real life' (1926&, pp. 79, 120—my italics). Analogously, as
he repeatedly claimed, the 'ill-omened kinship nomenclatures' (1932a, p. xx) only made sense
when seen in the contexts of speech, act and feeling, that is, the context of 'reality'. This
dichotomy between the 'ideal' and the 'real', between native 'theory' and 'practice', has,
overtly or implicitly, played a large part in the work of Malinowski's followers and successors. It
has stimulated ethnographic discoveries of very great importance, but has also given rise to
some theoretical misunderstandings. For Malinowski himself it went back a long way in his
thinking. It is foreshadowed in The Family among the Australian Aborigines (1913a) both in the
kind of questions asked about the 'actual' conduct of husbands and wives or parents and
children, to one another, and in the method of getting at the facts by confronting ethnologists'
interpretations with statements about the actual habits and practices of Australian aborigines
made by travellers and other observers. Here, for example, is the list of questions he wants
answered in order to discuss the relations of parents and children: 'Is there between parents
and children any kind of affection? What is the general character of the treatment of children
by parents? Are rudiments of education given by father or mother to offspring? . . .' etc. Verbal
statements by missionaries or even by native informants cannot answer such questions. Only
direct observation in the field can do so. My point is not that Malinowski insisted on field
observation as the only reliable source of ethnographic data. Haddon, Seligman and others had
proclaimed this before Malinowski went to the Trobriand islands and, indeed, it was because of
this belief that Seligman found the ways and means for Malinowski's field-work. What is
significant is the emphasis on practice (the activity; the behaviour; the concrete mutual
services; the exhibited self-interest, ambition, and vanity; the facts of mother love and paternal
affection; in short the actions and feelings and thoughts of individuals in social situations, as
directly observed by the ethnographer and as admitted by the actors) as the 'reality' of social
life, as agamst 'ideal', or Hheory\ the merely verbal formulation. Malinowski was, of course,
not alone in this way of thinking. The precursors of modern billiard-ball sociology^ held similar
views, only in more extreme forms. But his vigorous and brilliant propagation of the idea had a
direct influence on anthropologists and that is what concerns us here. In Mahnowski's case it
arose in part from the conditions of his field-work in the Trobriands; but it gained inspiration
also from his interest in and study of the new psychologies of the nineteentwenties. First
Shand's theory of the sentiments, then psycho-analysis, and finally behaviourism^ were the
dominant influences. This is easy to understand. For what these theories stressed in common
was the supremacy of the emotional and instinctive dispositions and tendencies in the human
make-up and the unreliability, if not impossibility, of knowing them through introspection and
reasoning. Psycho-analysis (to which he was introduced by Seligman) was the critical influence
on Malinowski's thought after his return from the Trobriands. Now the most obvious lesson of
psycho-analysis is that there is a large hiatus between the declared (conscious) motives and
affections of men and their deep-lying (unconscious) 'real* wishes and feelings. Furthermore, it
shows that antagonistic feelings can co-exist in the same person though he may not be
conscious of this. Malinowski's stock dramatization of the conflict between 'the main principle
of law, Mother-right, and one of the strongest sentiments, paternal love', is the translation into
ethnographic description of the Freudian concept of ambivalence. The 'reality of actual life' in
which the father is attached to his son rather than to his nephew, is opposed to the 'legal'
status of being not his son's kinsman—almost as if 'legal' here was identical with the conscious
fictions with which repressed motives are covered up, according to psycho-analysis.
Malinowski no doubt wrote in this way partly for effect. Yet his in- sistence on the test of
'reality' was a salutary rule in his pupils' field- work. It demanded a thorough mastery of the
native language; and this needed time and patience. The excellence of British ethnographic
field- work since Malinowski set the standard and laid down the methods is due chiefly to his
insistence on 'concrete' data. And yet it is by trying to satisfy this demand that the sociological
untenability of Malinowski's dichotomy has been made apparent. In almost every publication
on the Trobriand Islanders from Baloma to Coral Gardens, Malinowski gave notice of a
forthcoming book on kinship, several times significantly announced with the title of the
PSYCHOLOGY OF KINSHIP (e.g. 1932a, p. 434). Why was this book never completed? It may be
because Malinowski had said all he had to say about kinship in other pubUcations and had
nothing new to add. None of the major Trobriand works covers every aspect of kinship, in the
way, say, that Firth's We, The Tikopia (1936) does. But kinship figures in every one of them, and
not merely incidentally. It forms the true, though much concealed, framework of all except The
Argonauts (1922a). This alone would provide a substantial first-hand documentation on
Malinowski's theories of kinship. But in addition we have the series of articles on the family
and kinship'^ in which his ideas are explicitly and occasionally aggressively presented; and it is
clear from them that his theoretical views on kinship remained basically the same from 1922
till the end. But I do not believe that Malinowski would have been restrained from bringing out
his book on kinship by the thought that he had nothing new to add. He was, in fact, rather
given to repetition, the same hypotheses and even the same handful of dramatic cases being
used over and over again. I suggest that Malinowski could not write this book because his
theoretical premises ran counter to those on which any analytical study of a kinship system or
kinship in general must be based. In other words, the obstacle was Malinowski's concept of
kinship as 'the facts of sexuality, marriage, family and clanship' inter-related in 'one integral
institution—the Procreative Institutionof Mankind'. ^ The prerequisites for an analytic study of
kinship, that is, one which treats of kinship in its own right, have been familiar for thirty years.
Indeed they go back to Rivers and Morgan. They require the isolation of kinship as
'genealogical relationship recognised for social purposes' to quote the simple formula used by
Radcliffe-Brown (1929, p. 50). This means seeing kinship as part of the total social organization
or social structure which holds a society together. It means defining social organization as a
system of social relations, arising out of the modes of ordering and arranging the members of a
society in all their activities and therefore common to—cutting across—the 'procreative',
'nutritional', 'recreational', 'educational', etc., etc., institutions to which Malinowski gave the
name of functional aspects. These premises are the very opposite of those on which
Malinowski's need-centred concept of a Procreative Institution parallel to a Nutritional
Institution or an Educational Institution rest. The theoretical divergence, epitomized in the
conceptual distinction we now make between the social structure and the total body of tradi-
tional custom and usage to which the label of culture is usually attached, was once brought
home to me by Malinowski himself. He was discussing with me my field-work plans and we got
on to the subject of totemism. Comparing his own views with those of RadcUffe-Brown, he
said, 'The difference is that I am a functionaUst, and Radcliffe-Brown is a structuralist.' Better
evidence of his attitude is the tart, if somewhat obHque, rebuff in his Preface to Raymond
Firth's We, The Tikopia, to the criti- cism of the 'biographical method' impHed in Firth's
approach to kinship. Firth's main point (pp. 1 17-21), demonstrated by his meticulous
examination of Tikopia social life department by department, is that there is a common
'institutional map', a 'kinship structure' that can be extracted and seen as the common
'articulating' framework for all the different 'contextual situations' of residence, farming,
fishing, marriage, ritual, etc. The 'biographical approach', he says, is concerned with a special
problem that presupposes a kinship system. Hence the 'bio- graphical' explanation of the
'extension' of kinship terminologies outside the individual family is not tenable; for kinship
terms are not merely metaphors indicative of extensions of sentiment but correspond to
'codified obligations' and to modes of 'social regulation' as seen, for instance, in the
'representative status' of members of units that are 'fundamental to the particular social
structure' (p. 268). This is an analytical theory of kinship. 'Kinship behaviour and not kinship
sentiment is the study of the anthropologist', says Firth (p. 576), thus explicitly rejecting a
psychological starting-point. In parenthesis, he also rejects the interpretation of social
institutions, 'in terms of basic human needs'. No doubt that is why Malinowski makes a point of
restating his main convictions in the Preface. I have referred to Firth's book because it was the
first to confront Malinowski with field data fully measuring up to his own standards and aimed
at testing his kinship theories. It helped, also, to make clear to those of Malinowski's pupils
who were just embarking on research in the field the meaning of social organization as the
analytically distinct framework of all customary action and belief, in contrast to Malinowski's
concept of social organization as merely the 'personnel' of an institution tied to a 'material
apparatus' used in fulfilling basic needs in accordance with a 'charter' and 'rules'.^ In pointing
out the blind spots in Malinowski's view of kinship institutions, I want to make it clear that
there were good reasons for them in the beginning. As RadclifFe-Brown had said (1929, p. 53),
there was then a 'really important conflict . . . between conjectural history on the one side and
the functional study of society on the other'. Functional ethnography, in this context, had to
emphasize the descriptive reality of primitive social life, as for instance in the observable
concatenation of sexual motives and customs with courtship, marriage, family arrangements,
local grouping, work, magic and chiefship to which Malinowski devoted his superb gifts of
exposition. Malinowski's interest in psychological theories fitted naturally into his descriptive
aims and corresponded to his intuitive understanding of individual motive and attitude.
Functional theory as developed by him is an impressive attempt to provide v^^hat he would
have called a 'charter' and a systematic procedure for the kind of field research and
ethnographic reporting of which he was the unrivalled master. No better technique has yet
been devised for monographic studies, and we, Malinowski's pupils, have done no more than
follow the patterns he established. But Malinowski's conceptual scheme does not give to social
organization a generalized character and a systematic coherence of a different order from that
of custom, habit and motive. The facts of social relationship and social grouping are, in this
scheme, merely facts of custom and motive broadly on a par with, for instance, magical beliefs,
and springing in the last resort from such universal human instincts as those of parenthood or
such common human senti- ments as vanity and ambition. So we have nowhere a connected
analysis of Trobriand local organization, kinship, and political structure as an analytically
separable framework of social life, though the facts, or at least the bulk of them, are given
piecemeal where they are needed as the indispensable bones and joints on which to hang the
flesh and blood of his descriptions. This justifies my guess as to what stopped him from
finishing the promised book on kinship. In sum, what is inadequately stressed by Malinowski is
that kinship relations have to be seen as a system, within the framework of the total social
structure. Their fundamentally juridical nature then emerges, as Rivers appreciated. This is
specially relevant to the Trobriands where rank and residence rules directly affect conduct in
kinship. Questions of right and duty are, however, secondary to emotion and sentiment in
Malinowski's analysis of these data. This follows from his psychological approach; it is
consistent with the opposition he made between 'ideals' and 'laws' on the one hand and the
'reality' of action and motive on the other. 'The unity of the clan', he says (19266, p. 1 19), after
a vivid summary of the expression of this unity in rules of vendetta and mourning, exogamy
and food distribution, 'is a legal fiction in that it demands ... an absolute subordination of all
other interests and ties to the claims of clan solidarity, while, in fact, this solidarity is almost
constantly sinned against and practically non-existent in the daily run of ordinary life'. It is not
only a question of individual loves, hates and self-interest undermining 'the law' but of a
mysterious force called 'customary usage' which 'seconds' the free and aff"ectionate generosity
of father to son and thus 'defies' and 'circumvents' the 'rigid matrilineal law' of avuncular
obligation to the sister's son (cf. 1935, Vol. I, p. 205). The same kind of inference is drawn in
connection with every facet of kinship that appears in the Trobriand works. And the important
thing is that it is the alleged defiance of solidarity, the purported circumvention of the law that
is held up as being more 'real', certainly more authentically due to kinship, than the formal
rules. If this were not a matter of theoretical principle with Malinowski one would almost be
inclined to accuse him of suggestio falsi in these statements for, as we shall see, it is easy to
show from his own field material that his emphasis is distorted. What he is doing is to turn the
facts inside out, so to speak. This is most obvious in his handling of the odious subject of
kinship nomenclatures, to which I will return later. If it is easy to criticize Malinowski in
retrospect, it is just because he towers so high. The faults and failings, which we have learnt to
avoid by studying him, loom far larger than in lesser men. To criticize his treatment of kinship is
not to deny the brilliance and originality of some of his discoveries in this field, or to disparage
the inspiration his work gave to others. Indeed, I would maintain that Malinowski's most
productive hypotheses are concerned with problems of kinship and social organization in the
jural sense, and not, as he claimed, with questions of motive and of the meaning of custom in
the psychological sense. This may seem inconsistent with my criticism of him but it is not really
so. We must remember that his training in sociology was on the literature of Australian family
and kinship organization and that the Trobriand discoveries which most startled the
anthropological world were his revelations of matrilineal family relations in action. Even
without his initial interest he would have found that it was impossible to present his descriptive
data in an orderly way without considerable attention to the kinship framework. Among these
productive ideas I would include: the analysis of the sociological consequences of native
notions of procreation; the concept of the Initial Situation; the Principle of Legitimacy; the
theory of the incest taboo; and the concept of the 'splitting' of the paternal roles between
mother's brother and father. These are due, in large part, to the stimulus of psycho-analysis as
Evans-Pritchard pointed out in an appreciative article (1929). But it is very interesting to see
that Malinowski's response to psycho-analysis had roots going back to his Australian book.
Almost all the kinship problems to which he repeatedly returned in his Trobriand books, and his
theoretical articles, too, are foreshadowed there. He himself drew attention to this in one of his
most pungent and concentrated early statements of his kinship theory (19306). Kinship, for
Malinowski, all through his career, was not merely focused in the 'individual family'; it was
nothing but the individual family, considered either from within, or as the source of the
'extensions' that in his theory account wholly for extra-familial genealogically ordered
relationships. This point of view was firmly established in the AustraUan book. In a review,
which described the book as 'by far the best example in EngUsh of scientific method dealing
with descriptions of the customs and institutions of a savage people', Radcliffe-Brown (1914)
drew attention to the consequences of subordinating everything to the family. 'The Australian
notions relating to kinship', he said, 'cannot be studied vv^ithout reference to vi^hat the
author calls "group relationships" ... the relationship systems, classes, and clans. , . .' This
difference of orientation at that date was, it seems, prophetic, if we think of the later contrast
between Malinowski's notion of the Procreative Institution and Radcliffe-Brown's concept of a
Kinship System. As we know, at the very time that Malinowski was working on his vindication of
the individual Australian family as the 'basis of their social structure', Radcliffe Brown (1913)
was in the field investigating those aspects of Australian kinship institutions that Malinowski
considered to be of secondary importance. Malinowski's book demolished the theories of
primitive promiscuity, group marriage, and clan priority. But far more important was his
contention that all the economic, social, legal, and ritual customs and practices of the
Australians converge on the nodal point of the family organization; and, partly because it began
as a move in opposition to group marriage, etc., the emphasis is on the 'actual working' of the
family in terms of the individual relationships of its members. Instead of discoursing on
marriage classes and clans, he deals at length with the economic, sexual, and affectional
relationships of spouses, insisting particularly on the 'individual appropriation' of the wife by
the husband. He soon comes up against the controversy that was destined to be one of his
main theoretical preoccupations thereafter. This is the problem arising out of Spencer and
Gillen's researches, of whether or not the Australians have a knowledge of the physiological
function of the male in procreation. After minutely studying the evidence he concludes that
they do not, over most of Australia, and he argues that this is as might be expected from their
general ignorance of natural science. This leads him to expound what later became the
important distinction between physiological paternity and social father- hood. The reader gets
a foretaste of the later Malinowski at his best in the analysis of the sociological significance of
incarnation beliefs. These are shown to imply a pre-existing kinship relation of every child with
its parents, and to fit in with ignorance of the physiology of reproduction. 'Consanguinity (as a
sociological concept)', he says, 'is therefore not the physiological bond of common blood; it is
the social acknowledgement and interpretation of it' (1913a, p. 182). When, following on this,
Malinowski tries to ascertain just what the 'emotional character of the parental kinship
relationship' (1913a, p. 199) is and what are the actual procedures and tasks of feeding, caring
for and bringing up a family, we have the ground laid for the Principle of Legitimacy, the Initial
Situation and the working out of the theory of incest. It is worth observing how the same
interests run through The Natives of Mailu (1915a). Malinowski made a special point of
investigating the sexual customs, the beliefs about conception, and the individual family
relationships of the natives. He finds that the Mailu are ignorant of the nature of physiological
paternity. His primary interest is in the 'reality' of action, behaviour, sentiment, and conduct.
This is nicely shown in the discussion of legal institutions (1915a, chap. Ill, sect. 5, pp. 576 ff.).
The nucleus of Crime and Custom (19266), with its insistence on the division between 'ideal'
and 'real', is here. The ethnographer's task is defined at length. 'Actual, concrete facts' are what
he must get; and Malinowski complains that he was unable to get 'authentic stories about
murder, adultery, theft' and other crimes in the old days. He asserts that 'in every community
there exist fundamental rules which must be observed. The infraction of these rules is a lurking
temptation and there are always individuals who succumb. . . . To discover the rules, the
possibilities of infringement, the restraining forces . , .' is his aim. And after some discussion he
expresses his conclusion that 'the conception of criminal law or of civil law, or of the distinction
between the two' has no 'counterpart in native ideas'. I am giving prominence to the continuity
between Malinowski's theoretical training, his first experience of field-work and his mature
ideas and theories because it seems to me that this helps to explain why he, more than any
other anthropologist of his time, responded so eagerly and creatively to the influence of
psycho-analysis. This may seem an exaggeration if we think of the impact of psycho-analysis on
the leading anthropologists of the nineteen-twenties—for example, on Rivers and both the
Seligmans, in this country, and on Sapir and Benedict in America. Even Radcliff"e-Brown, later
so rigorous in his sociological thinking, must have been influenced by Freud, for his theory of
joking relationships obviously owed something to Freud's theory of wit. But for Malinowski
psycho-analysis was the light by which he was able to make a new synthesis of his ideas and
experiences. It is important to understand clearly what Malinowski's innovations were. For, of
course, ethnological field-workers and scholars usually included accounts of sexual customs
and practices, beliefs about the nature of conception, and customary modes of conduct
between parents and children and between other kinsfolk, in describing primitive kinship
institutions. The nineteenth-century classics are full of such discussions. Was not the
controversy about primitive promiscuity and group marriage chiefly concerned with the
question of the connection between sex and parenthood? More pertinently, the pioneers of
scientific field-work had paid considerable attention to these matters. Thus Rivers provoked
Malinowski's scorn by his confused speculations about the origins of such customs as the Banks
islanders' avoidance of affinal kin in earlier conditions of 'sexual communism' (e.g. Rivers, 1914,
Vol. n, pp. 136 ff.). But these speculations, related to Rivers's preoccupation with the problem
of the mother's brother, were associated with shrewd observations similar to those on which
MaHnowski later built his analysis of the position of the father in the Trobriands, Rivers noted,
for example, that a person was not considered to be his father's kin in the Banks islands
because he belonged to the opposite moiety. Malinowski's break with this kind of ethnology
lay, therefore, not in his choice of topics but in his handling of them, especially after his field
experience in the Trobriands. This was first shown in 'Baloma' (1916), where his 'contextual' or
'institutional' method led him to make a particular point of what he called the 'social
dimension'. Individuals and groups in fact vary greatly in their observance of custom and in
their knowledge of the beliefs and practices which we conventionally attribute to a tribe as if
they were uniformly present among all its members. Hence, he argued, the actual language
and gestures used, the feelings, motives and ambitions of actors and informants, the whole
psychological setting, are important data for ethnography. Custom and belief thus appear as
the motives of action, both in the positive sense of furnishing aims and incentives, as in the
kula or gardening, and in the negative sense of provoking instinctive rebellion, as in the evasion
of incest taboos. To seek for psychological explanations of the springs of conduct came
naturally to MaHnowski and was consistent with his rejection of both historical theories and
collectivist sociologies of a Durkheimian type. In common with most scholars in the social
sciences at that time, Malinowski was an admirer of McDougall and Shand, Psycho-analysis
could be regarded as taking their theories a step further and this made it the more acceptable
to him. Thus, psycho-analysis fitted Malinowski's general approach. It also fitted his conception
of human nature. We see this from the following statement of what it was in psycho-analysis
that appealed to him: 'The open treatment of sex and of various shameful meannesses and
vanities in man—the very thing for which psycho-analysis is most hated and reviled . . . should
endear psycho-analysis above all to the student of man.' 'Man is an animal', he continues, 'and,
as such, at times unclean, and the honest anthropologist has to face this fact' (1927a, p. viii).
This unflattering view of man was in the spirit of the nineteentwenties; but it was not just a
case of drifting with the tide. Malinowski played a lively part in creating this climate of thought,
more particularly with his writings on sex and the family. Did not the Trobrianders become a
byword among novelists and sex-reformers, as well as the model primitive society (I am
tempted to say, the Twentieth-Century Noble Savage) for textbook reference among the most
diverse of scholars and scientists? What better testimony could there be to the authenticity of
Malinowski's understanding of man? To him, as to Freud, the seamy side of human conduct
and belief was a matter of observation not of fashionable ideology. To sum up, Malinowski's
debt to psycho-analysis is obvious in much of his work (e.g. on magic) but nowhere so
markedly as in his descriptions of Trobriand family structure and in the theory of kinship that
grew out of them. Indeed I would maintain that it was the notion of the (Edipus Complex that
gave Malinowski the inspiration for the main features of his kinship theory. Two features of
Freudian theory specially impressed Malinowski. The first was the picture of the ambivalent
emotional relationship of father and son arising out of the clash between the son's instinctive
sexual urges and the culturally enforced power and authority of the father. Duly transposed
into 'matriarchal' terms this gave coherence and meaning to the ostensibly contradictory facts
of Trobriand family relationships. Here was a mechanism which accounted neatly for the
coexistence, one might almost say the necessary coexistence, in a system of mutually opposing
forces, of the distinctive elements of the matrilineal family. Father love and avuncular
authority, the sexual bonds of spouses and the sexual barriers between brother and sister, the
obligatory friendship and common interests of uncle and nephew alongside of their mutual
enmity ('in real life to a certain degree and quite openly in myth'—1927a, p. 121), the
ignorance of paternity, the incest rule and the urge to break it, all fall into a unitary and
consistent scheme on the revised Freudian hypothesis. It accorded well with Malinowski's way
of thinking about 'real' emotions and sentiments as being in perpetual struggle against 'ideal'
laws and morals. The hypothesis had another merit. It was economical and realistic. Given the
irreducible biological facts of sex and of child nurture, everything else followed from the
necessary constitution of the in- dividual family. No extraneous data or assumptions were
needed. And the conflict between sexual wishes on the one hand, and the demands of law and
morals on the other, which was deemed to give rise to the CEdipal ambivalence, could, it
seemed, be confirmed by direct observation in the field. This brings up the second aspect of
psycho-analytic theory which specially impressed Malinowski. The (Edipus mechanism seemed
to be rigorously accounted for by a kind of causal process. It was the result of a genetic
sequence beginning with inborn infantile sexuality and running through all the stages of child
development in an interplay of innate impulse and the 'moulding' influence of cultural training.
The whole process was, Malinowski thought, verifiable by field observation (i.e. 'real') and had
the support of general psychological theory. This genetic conceptualization became and
remained the first principle of Malinowski's kinship theory even when, later on, he turned
against psycho-analysis. The concept of the Initial Situation and the notion of kinship extension
arise directly from this principle. He spoke of it as 'doing, somewhat tardily, for social
anthropology what psychology has been doing for the study of the mental development of the
individual', as the study of the 'moulding of innate dispositions' ( 1 930a). He summed up the
method of approach in such formulas as 'It is, therefore, the process of the extension of kinship
from its extremely simple be- ginnings in plain parenthood [sc. the Initial Situation] to its
manifold ramifications and complexities , . . which . . . forms the real subject- matter of the
study of kinship' (19306—my italics). As I have said, a psychological framework was essential to
Malinowski's functionalism. Everything he wrote was riddled with psychological explanation
partly because his functionalism meant seeing custom as motive, partly because its
instrumental and utilitarian form led back to physiological needs, and the simplest way in
which these can be visualized as emerging in action is as the driving forces behind instincts,
sentiments and emotions. A graphic instance is Malinowski's theory of magic, where, it will be
remembered, he uses the notion (originated by Freud and adopted, in a watered-down
physiological form, by Rivers) of 'spontaneous ritual and verbiage of overflowing passion or
desire' related to 'natural responses' of fear and hope and 'illusions of subjective experience'
(1925a, pp. 75-6). But necessary as psychological theories are for it, a genetical process need
not, strictly speaking, be posited in order to introduce concepts of motive and feeling into
ethnographical description. Every ethnographer uses such concepts. Malinowski himself does
not give a genetical derivation for the psychological concepts he uses in describing breaches of
the incest taboo or the conduct and feelings of the Trobriand husband to his wife and father to
his children. For these, like most ethnographic descriptions, are set in the generalized present.
The truth is that the concept of genetic development was less a psychological hypothesis than
a methodological one for Malinowski. It was the functionalist answer to 'imaginary
speculations' about kinship, in fact a sounder theory of origins and of temporal sequence
because ostensibly verifiable by observation. It could be seen happening in the reality of social
life. The origins of the clan, said Malinowski, scoffing at these speculations, 'happen . . . under
our very eyes ... I have myself witnessed (them) . . .' (19306). And what he puts forward is his
hypothesis of the 'biological development of kinship', which leads to the conclusion that 'the
clan develops as a derived sociological form of grouping by empirical processes . . , along the
life history of the individual'. Malinowski disliked the systematic statement of theory. He even
made a virtue of using technical terms loosely and was criticized by Radcliffe-Brown (1929) for
confusing the concept of descent with that of bilateral kinship. As he had a flair for coining
valuable new terms this was exasperating. But inconvenient as this may be to his readers,
avoid- ance of system and precision suited his habits of thought. For his insight worked in
leaps, like stepping stones in the flood of his divagations. Thus the 'life history' principle was
ideal for his mode of thought. It was naturalistic enough to fall in with the 'institutional' form of
presenting his descriptions and arguments, and yet flexible enough to give him plenty of room
to spread about without losing the semblance of an objective order. Most of all, it enabled
persons to be shown as active agents in the process of developing sentiments, dispositions,
habits and ideas. Thus the extrapolation from the 'initial situation' in the parental family to
descent concepts, clans, and tribal structure is effected by successive genetical steps. The
logical flaws and the empirical selectiveness that are obvious when these facts are looked at in
terms of their con- temporary relations in a total social structure, are easily concealed in such a
piecemeal presentation. I want to underline this methodological value of the genetic concept
for Malinowski. For, in spite of his protestations, he never himself made any true 'life history'
studies. What he wrote under that rubric are highly generalized and abstract statements of
sequences of customary activities and patterns of relationship in a model matrilineal family and
clan system. Even in his ethnographic work, as for instance The Sexual Life of Savages, where
he follows a rough order of chronological succession in describing the vicissitudes of the sexual
instincts and proclivities of the Trobrianders, there are no individual 'life histories'. The 'bio-
graphical' method gives only a convenient framework on which to hang the descriptive facts of
kinship, as Firth and others after him have found. But it would be a mistake to think that what
Malinowski got from psycho-analysis was only a model for a procedure of description. He also,
as I have noted, got the clues towards some of his most cherished and valuable theoretical
ideas on the nature of family relationships. For this he had to reinterpret Freudian theory to
accommodate his ethnographic insight. The main problem was the empirical status of the
unconscious emotional formations and conflicts, overtly visible only in disguised or symbolic
forms, upon which the hypothesis of the CEdipus Complex relies, and how they couldbe
compared with the observed facts of customary and publicly accepted behaviour which are the
ethnographer's data. Freud himself, incidentally, was well aware of the problem. In Totem und
Tabu he frequently comments on the danger of identifying the symptoms and mental states of
neuroses with primitive customs. He emphasizes that it is analogies and coincidences not
identity that he is investigating. Malinowski at first followed Freudian terminology and claimed
that his observations supported Freudian theory (e.g. 1927a, pp. 82, 92, etc.). He spoke of the
'repressed passions' breaking through, of 'the repressed sexual attitude of incestuous
temptation . . . towards the sister', and of the 'ambivalent attitude' of sister's son to mother's
brother (p. 80). But the controversy with Ernest Jones, and criticism of the hypothesis of a
primordial parricide, led him to formulate his notion of the 'nuclear family complex' as 'a
functional formation dependent upon the structure and upon the culture of a society' and
'determined' by 'the manner in which sexual restrictions are moulded . . . and authority
apportioned' (p. 143). This was Freud rephrased but with the stress on the primacy of culture
over instinct in the formation of the nuclear complex. Instead of a parallelism between the
customary attitudes and behaviour of Trobrianders and the partly unconscious emotional
formations revealed by psycho-analysis, Malinowski was led to assert that custom is in fact the
equivalent of unconscious emotional forces and to reject the need for the concept of
unconscious motive (pp. 173 ff.). The 'nuclear family complex' became 'a configuration of
sentiments typical in a patriarchal or a matriarchal society' (p. 178) and by definition directly
accessible to ethnographic observation. This was a significant discovery and has been tacitly, if
not openly, accepted by ethnographers since then. Malinowski was thus confronted with the
task of showing how culture regulates the elements of family organization, 'apportions'
authority, and 'moulds' family sentiments. He found part of the answer in the hypothesis of the
'rule of legitimacy' which, he said, reinforces the instinctive parental feelings of the human
male and gives cultural definition and legitimacy to motherhood as well as to fatherhood. It is
connected with the 'two main perils of humanity: the tendency to incest and the revolt against
authority'. It was in establishing his thesis about the nuclear complex that Malinowski arrived
at his remarkable reinterpretation of the Freudian theory of the incest taboo. The stroke of
sociological inspiration was to relate the incest taboo to the function of the family as the
agency through which the knowledge and sentiments essential for maintaining culture are
transmitted from generation to generation. This educational task requires emotions of
'reverence, dependence, respect as well as strong attachment' to the mother and 'thorough
submission to the leadership' of the father. So if, as psychoanalysis holds, incest is a universal
temptation it must be forbidden because it is incompatible with the existence of the family as
the medium through which culture is transmitted. If incest were permitted 'the fundamental
pattern of all social bonds, the normal relation of the child to the mother and father would be
destroyed' (1927a, p. 252). To this general hypothesis he added a supplementary analysis of
the differences between 'patriarchal' and 'matriarchal' family and descent systems from the
point of view of incest prohibitions. 'Unilateral kinship' as he calls it (p. 268) is casually
disposed of as 'the only possible way of dealing with problems of transmission of possessions,
dignities, and social privileges'. Father-right is thus characterized in rephrased Freudian terms:
'In the patrilineal society the father has to incorporate in himself the two aspects ... of tender
friend and rigid guardian of law. This creates both a disharmony . . . within the family . . .
disturbing co-operation and . . . jealousies and rivalries at its very heart' (p. 270). Matrihny, on
the other hand, is represented as being better adjusted to the psychological facts. As it is the
mother's brother, not she herself, who wields 'coercive powers' this does not introduce
jealousies between her and her son. The sentiment between mother and child is not dis-
turbed by the transmission of legal and economic powers. The mother's brother 'who
represents stern authority, social ideals, and ambitions is very suitably kept at a distance
outside the family circle'. We note here that the emotional attitudes attributed to father,
mother and uncle are of a highly general kind and are presumed to be quite overt. One of the
paradoxes in Malinowski's work is that his theoretical writings are mostly couched in terms of
the highest generality, in strong contrast with the rich and concrete detail of his ethnographical
books. This is conspicuous in everything he wrote on the theory of kinship. The specific facts of
custom, with their variability from culture to culture, disappear behind abstract terms. Closely
looked at these terms —even quite conventional terms like marriage, paternity, maternity,
etc.—prove to be concerned with the limiting principles and factors of kinship rather than with
the realities of custom, belief and sentiment. Where they do not deal with physiological and
psychological principles that form necessary prerequisites for the functions of social and
cultural reproduction in human society, they deal with structural prerequisites. Such, for
instance, is the notion of the universal need for a husbandfather, in the first place as protector
of the mother and child, but also in order to confer legitimacy on the procreative role of the
mother. And what is more, all these principles and hypotheses are deductively reached. One of
Malinowski's chief debts to Freud was that psycho-analysis joined to his Trobriand field data
provided the premises for his most fertile deductive hypothesis. The theory of incest is an
outstanding instance. I do not think this is a purely tautologous generalization. For it does
specify definite rules of sexual behaviour that must be obeyed for the social and cultural tasks
that can only be performed in the relation of parenthood to be fulfilled. It is the relation of
parenthood, be it noted, that is central to Malinowski's hypothesis, not the composition of the
individual family, though he states it in language that implies an arrangement of co-resident
parents and children. This is important; for the generalization is susceptible of being refuted by
direct observation of the relations of parenthood, irrespective of residential arrangements or
even of the duration of marriage. If Malinowski's contribution to kinship theory on the
conceptual side was, in effect, a series of loosely connected, largely deductive hypotheses
about family structure, v^^hat was it on the ethnographic side? This is important because
Malinowski was at his best in the descriptive exploration of an idea through the medium of his
own field observations, and his influence was chiefly exerted through his ethnographic
writings. As far as the study of kinship systems is concerned, I think his influence has been
greatest through the models he has provided for field research and for the descriptive
presentation of kinship custom and behaviour. The genetic framework or, more appropriately,
the life- cycle sequence is well suited for this and the insistence on observing and recording
what takes place in the living situation is fundamental. More specific models are provided by
almost all of his books and it would be easy to give a list of ethnographic publications that have
been inspired by them. No ethnographer with a 'functionalist' training writing about kinship
today will overlook the rivalry and conflict aspect of the relations between the legal possessor
of parental authority and the filial generation, or the significance of reciprocity rules in kinship,
or the educational (sc. 'socialization') function of the family and its connection with incest
prohibitions. One of the best examples of such a model is the essay on The Father in Primitive
Psychology (igzjb). Though much of the ground it covers is the same as in Sex and Repression
(1927a) and in later works, this essay brings into particularly clear relief both Malinowski's
ethnographic method and his most influential contribution to kinship theory. This was the now
classical picture, to which I have already several times referred, of matrilineal paternity divided
between the father, with his personal bonds of aff"ection and solicitude for his child, based on
his exclusive sexual rights over the mother and his part in the task of childrearing, and the
mother's brother with his bonds of right and duty with his sister's son, rooted in the incest
barrier, in legal authority and in descent, and accompanied* by suppressed hostility.
Malinowski found the clue for this analysis in the Freudian concept of ambivalence in the
(Edipal relationship of father and son but he had the insight to apply it to the structural
arrangements of the 'matrilineal family', and his method showed him that the emotional
pattern was supported by the physiological ideas, the ritual beliefs, the conventional loyalties
and hostilities, the residence patterns and the legal rules of the society. As a demonstration of
the significance of functional consistencies in social organization it is noteworthy, and has
inspired many confirmatory studies, like those of Richards (1934, 1950a),
Fortes(i95o)andNadel(i95o). Butt hiswas not MaUnowski'sownjudge ment of his analysis. For
him the issue was one of a dramatic conflict between passion and law which was undermining
the family at its core. - He saw it in terms of '. . , beliefs, ideas and customary rules which
smuggle extreme patrilineal principles into the stronghold of mother right' (igzy^*, p. 85). This
was not just rhetoric or an example of Malinowski's ingrained contempt for precise analytical
concepts. He held, as a point of theory, that what he was describing was 'patrilineal principles'
in conflict with 'matriliny'. He adhered to this conception throughout his work. Even in Coral
Gardens (1935) where there are indications of appreciation of the analytical autonomy of social
organization in its jural aspect, this point of view still prevails. Its significance for later research
is that it treats the analytically heterogeneous topics of sex and sentiment in the setting of
dyadic personal relationships, on the one hand, and the legal and political order of the total
society, on the other, as if they were all on one level. This may be justifiable in the context of
the parental family taken as a more or less closed system where jural and political status are
subsidiary to the inter-individual relationship though even in this context they are not
negligible, as Malinowski himself showed. It breaks down when the family is considered in the
external context of local, jural and political relations. The proof lies in Malinowski's failure to
understand the nature of classificatory kinship terminologies; for to give Rivers his due,
however wrong his explanations were his empirical insight was sound. These terminologies are
not just metaphors and homonyms as Malinowski proclaimed. They are indicators of social
relations and of modes of grouping and arranging persons which are present in the social
structure of any people at the same time as family relations and which serve to bind the family
into the total jural and political order. The notion put forward quite early (e.g. 1927a, p. 221)
and adhered to all his life that *in all primitive societies, without exception, the local
community, the clan or the tribe, is organized by a gradual extension of family ties' needs only
to be stated for its absurdity to be seen. It led to entanglements such as the contradictory
assertions that the 'clan always grows out of the family, forming round one of the two parents .
. .' and three lines later, 'family and clan diflFer thus profoundly in origins, in the functions
which they fulfil . . . etc' (19306). It led to the series of logical in- consistencies that culminated
in the curious hypothesis of 'strata' of kinship nomenclature in which the same sounds had
diflferent meanings because 'the individual, the extended, the local and the classificatory are
diff^erentiated by phonetic distinctions . . . circumlocutions . . . contextual indices' (loc. cit.),
these being 'real' as opposed to the 'spurious' nature of the terminologies in the verbal sense.
We can see that Malinowski's error was to confuse the frame of reference of the individual life
history—or rather of the conventional life history of the standard individual—with that of the
total social structure. It is excellently documented in his only effort to explain Trobriand kinship
terminology. In The Sexual Life of Savages (1929a, pp. 434-5) there is a table of Trobriand
kinship terms, and this is fol- lowed by some discussion of the terms. His approach is well
illustrated in the explanation of the term for sibling of opposite sex. It is linked to the incest
taboo. The term is given its meaning by the emotional conditioning of the child due to the
reprimands and horror of 'its elders' whenever it makes 'playful advances to the other small
being'. On the same principle, Malinowski soon arrives at this extraordinary fallacy: 'The
anomalous extension of the word for father (tama) to the father's sister's son is important, for
it demonstrates the influence which language has upon customs and ideas . . .' and so on. The
whole paragraph deserves study as an example of a beautiful methodological and logical
howler due to preconceived ideas; and this was no slip of the pen for the rest of the discussion
on Trobriand kinship terminology follows this line.^ What Malinowski refused to accept was
the fact that kinship terms designate jural relationships and groupings. The Trobriand kinship
system, we would say today, has an obvious 'Choctaw' type terminology, well exemplifying
principles formulated by Radcliffe- Brown at this time and published in 1931. The usages
tortuously made to look anomalous by Malinowski are perfectly normal for such systems and
enable one to infer certain jural consequences, such as the probability of property rights being
vested in a matrilineal descent group. But if we know this now, could Malinowski have
anticipated these results of recent theoretical research? We must, I think, answer that he could
have avoided the blunder, if he had made more of an effort to understand the findings and
interpretations of other students of kinship. He might, indeed, have done this if he had
adhered to a rule of method reiterated in The Family Among the Australian Aborigines, that it
is specially important to bear in mind the connection of family organization 'with the general
structure of society' (1913a, p. 300). What he in fact did was to revive, in a new-fangled form,
Kroeber's early view^ of kinship terminologies as primarily linguistic and psychological facts,
not 'determined' by 'social conditions'. He knew very well, from the investigations of Seligman
(1910) and Rivers (1914), that the alleged idiosyncrasies of terminology he was discussing have
a wide distribution in Melanesia; and he might have asked himself whether it was not as
ludicrous to invoke linguistic anomalies to explain such general features of social organization
as to refer them to antecedent marriage rules. He might have given serious consideration to
some of Rivers's ideas—not to mention earlier writers such as Kohler, and contemporary
scholars like Radcliffe-Brown and Lowie. Thus he might have followed up Rivers's contention
(1913, p. 72) which Lowie adopted (1920, chap. VI)—but stripped of its pseudo-history—that
certain types of kinship nomenclature are correlated with exogamous clan systems, for he
himself remarked that one of the 'proofs' of Trobriand 'clan soUdarity' '. . . is the classificatory
use of kinship terms' (1929a, p. 425). Malinowski's unwillingness to see the meaning of kinship
terminologies in the setting of the total social structure did not, it should be added, pass
without criticism at the time. A good instance is Mrs. Seligman's paper on incest and descent
(1929). While accepting Malinowski's hypothesis of the relationship between incest
prohibitions and family organization and agreeing with him about the central importance of
the family in the transmission of culture, she insists on the 'equal importance' of a 'larger
group' in which a 'system of descent' is the essential element. Though not quite clear in her
remarks about classificatory kinship terminologies, she sees that 'the acceptance of the
classificatory system of relationships carries with it the status of those relationships, not its
emotional equation'. Again, Malinowski never succeeded in persuading his pupils to follow him
wholeheartedly in this matter. Thus, without directly debating the point, Raymond Firth makes
it quite clear, in his elaborate analysis of Tikopia kinship language (1936, chap. VII), that he
does not accept Malinowski's ideas on this subject. His argument is eclectic but he underlines
the fact that kinship terminologies serve as indices of jural groupings and classi- fications, such
as the attribution of 'representative status'. There is one particularly telling example of how
Malinowski confused issues by lumping together indices of the jural order of society and
usages showing up individual motive and sentiment. We know how obsessed he was with the
dichotomy between 'ideals of tribal morality' and 'the real behaviour'. He maintains (1929a, p.
426) that 'the smoothness and uniformity, which the mere verbal statements suggest as the
only shape of human conduct, disappear with a better knowledge of cultural reality'. The thesis
was repeatedly illustrated by reference to breaches of the law of incest and exogamy. Now one
of Malinowski's most firmly held hypotheses was that the law of exogamy is simply an
extension to the clan of the prohibition of sexual intercourse between brother and sister or
parent and child which is incest in the strictest sense. This idea gained wide currency as an
established generalization. For though Malinowski did not originate the idea he, more than
anyone else, gave it the appearance of cogency by showing it in action. It was indispensable for
his views about the nature of 'cultural reality' and the basically linguistic and psychological
meaning of kinship terminology. But his own comments and descriptions afford ample
evidence that the Trobrianders, like many other peoples in whose social structure unilineal
descent groups enter into the regulation of marriage, do make a jural distinction between the
two rules; and not only that, it is also quite clear that they have severe and effective means of
enforcing the 'ideal' rules. It is simply not true—or at any rate, not proven by the one or two
cases quoted—that 'the breach of exogamy within the clan, intrigues v/ith . . . kindred-in-clan .
. . though officially forbidden, ruled to be improper, and surrounded by supernatural sanctions,
is yet every- where committed' (my italics). But the point I particularly want to bring out is best
put in the remark that follows immediately on the paragraph from which I have just quoted
(1929a; p. 431). ^Marriages—as distinct from intrigues within the clan are definitely regarded
as a serious breach of the rule' (Malinowski's italics); and, a bit further on, *. , . the rule of
exogamy, far from being uniform and wholesale in its application, works differently with regard
to marriage and to sexual intercourse . . .' Finally comes the information that incest with the
own sister or the mother's sister's daughter is 'a real crime . . , a dreadful crime' which may
lead to suicide. It is obvious that the Trobrianders do distinguish, in thought and in their
institutional arrangements, between exogamy, the rule prohibiting marriage 'within the clan',
and the incest prohibition which applies to sexual relations between close kin. Moreover,
exogamy is evidently enforced without exception, as we should expect with a jural obligation
that can only be entered into with the consent of the woman's kin and is validated by
prestations on both sides (cf, the description of the procedure, Malinowski, 1929a, chap. IV,
sect. 2 and 3). Again, the range of close kin subject to the incest taboo is equally clear. It
includes the members of the same sub-clan (dala) or, as we should probably now call it, local
Hneage,^ not only the closest circle of maternal kin; and within this range the breach of the
incest taboo is evidently so outrageous and rare that it evokes the extreme sanction of suicide.
The 'everjrwhere' in the earlier quotation seems to come down to occasional intrigues, but
definitely not marriage, between members of different local 'subclans' of the same clan. By the
test of regularity and incidence of observance, the 'cultural reality' of incest and exogamy rules
is just what the 'ideal' states it to be. This example well illustrates some of the weaknesses of
Malinowski's theory and method in the study of kinship. He would have avoided the error if he
had started with a different frame of analysis. Thus, a crucial problem is the exact local and
genealogical make-up of, and relationships between, the sub-clan and the clan. Taking all the
sources (and es- pecially 1935, Vol. I, chap, XII) together we can construct some kind of picture
of the sub-clan as the corporate, permanent, localized, land, rank, and magic-owning group
made up of ranked 'lineages' connected by assumed common descent; and of the clan as one
of four widely scattered, named divisions based on a totemic myth of remote common
matrilineal ancestry. But we can only guess what the implications of this significant fact are for
marriage rules and relationships. We can note, for instance, that the cases actually quoted of
so-called breach of exogamy seem to occur between members of different sub-clans of the
same clan (e.g. 1935, Vol. I, p. 385). In the context of a strict analysis of clan and sub-clan
structure, the law of exogamy would have been seen in its true quality as a jural institution.
This would have been equally clear if Malinowski had examined his data from the point of view
of the marriage contract and the rights and duties arising out of it for the spouses as married
partners and as parents. And there is yet another alternative. If (remembering his Durkheim)
he had bethought himself of the chronic anomie into which Trobriand society would fall if the
laws of incest and exogamy were commonly flouted, he would have seen the issue differently.
He would have realised that such a state of affairs would make nonsense of the Principle of
Legitimacy to which he rightly gave so much importance in the context of marriage and
parenthood. I have dwelt on this example partly because it concerns a subject which was
prominent in Malinowski's interests all his life and partly for a personal reason. During my first
year of field-work among the Tallensi I tried to confirm Malinowski's hypothesis in the field. It
was thus that I learnt to distinguish between the jural and the personal, or psychological,
components of kinship institutions.^ Moreover, it is an example which explains why there is
and always will be inspiration to be found in the careful scrutiny of everything Malinowski
wrote. This is because the quality and amount of his ethnographic observations, coupled with
his exacting, almost pernickety thoroughness in setting them down, give us a wealth of data for
checking his theories by direct comparison with the facts. Coral Gardetis and Their Magic is the
acid test of this. On the surface, kinship and social organization appear as incidental topics only.
But patient study reveals that here, more than in any of the other Trobriand books, they serve
as the hidden 'articulating principles' of the whole book. The reader has, of course, to work this
out for himself since the relevant data are scattered piecemeal through the narrative of
gardening activities. In doing so he will find that the data are there to check and sometimes to
modify and even to refute Malinowski's generalized state- ments. He will be helped also by
digressions on particular topics of social organization where Malinowski's insight breaks
through and he gives a structural analysis very close in form and even in idiom to the practice
of today. It may be that he was responding to a tendency that was already 'in the air' in 1932, A
good instance is the description of 'the structure of the sub-clan' in the chapter on Land
Tenure. To judge by his own remarks (1935, Vol. I, p. 481) Malinowski was specially satisfied
with this description. It is carried out in quite rigorous sociological and jural terms, the red
herring of 'real' versus 'ideal' in conduct being, for this task, left aside. Thus we are told that
'magic, strictly speaking, should always be handled by the eldest member of the senior lineage
in every sub-clan' and that 'sometimes ... it is not given to younger brother or maternal
nephew, but to a person who has no legal place in the lineage ... the headman's own son', who,
however, usually 'becomes naturalized by the act of cross-cousin marriage, which gives him
almost /m// legal right to reside in his father's community' (1935, Vol. I, p. 349—my itaUcs).
We learn that 'the right of citizenship . . . the absolute and unquestionable right of residence'
from which the right to cultivate soil derives, is gained only by membership of a sub-clan
which, if we patiently pursue the theme, we find has not only a 'local character' but also a
'spiritual unity and continuity' expressed in beliefs about the after-hfe and in the ownership of
magic (1935, Vol. I, pp. 344 ff. —my italics). There is much more in this vein in the book. I have
quoted these short extracts just to show the affinity of this mode of handling the data with
present-day structural theory, even in some items of terminology.^ Not that the mode of
thought followed in previous books is discarded. The narrative relies heavily on the attribution
of motive and imputation of sentiment. But descriptions and discussions of what Malinowski
called the 'juridical' aspects of social structure are more prominent and abundant here than in
the previous works, perhaps because of the subject matter, and the governing principles can be
extracted. As this is not an essay on Trobriand social structure but on Malinowski's ideas, I will
consider only one or two examples. They show both the brilliance of Malinowski's insight in
ethnographic particulars and his failure to follow up his analysis in a systematic way. An
excellent instance is the urigubu gift of agricultural produce annually handed over by men to
the husbands of their sisters. It appears in every description Malinowski gave of Trobriand
family organization. It is a major topic in Coral Gardens, where it comes up in one context after
another and is given a comprehensive chapter to itself. It was the descriptive features of this
picturesque custom that loomed foremost, from the first (e.g. 19220), in MaHnowski's
imagination and he therefore turned the HmeHght on to the apparent paradoxes in it. In Coral
Gardens it is presented from the outset as being 'perverse' and 'so difficult to grasp' that even
'long-time white residents . . . married to native women . . . who . . . benefit under the
Trobriand harvesting system . . .' cannot understand it. The problem thrust into the fore- front
is 'what motive can make one man offer the best part of his harvest to another?' And it is made
to look more portentous by long disquisitions on the 'terminological looseness', the numerous
'ex- ceptions', the paradox (to the ignorant European often dragged in as a lay figure by
Malinowski for rhetorical emphasis) of a man's giving away his choicest crops, the 'puzzle' as to
whether they are given to the sister's husband or herself or her family, the ceremonial
transport of the gift, the economic waste implied, the conflict between a man's 'interests and
his heart' which are fixed in his own household of wife and children, and his 'pride and moral
duties' which are 'in the household of his sister'. (These references are from 1935, Vol. I, chap.
VI, but they can be paralleled from the other Trobriand books.) In Coral Gardens this familiar
build-up is followed by something new. This is an analysis, in fairly strict structural terms, of the
overlap be- tween the 'unit of filiation' made up of brother, sister and sister's children, and the
'paternal household'. Marriage is defined as a 'publicly acknowledged relation, approved and
accepted by the girl's family, and binding them to definite economic prestation'. This includes
not only the reciprocal services of husband and wife and of father and child, but the urigubu;
and we now get a structural and juridical analysis of this gift which makes all the preceding
dramatization of it look irrelevant. 'The urigubu', Malinowski sums up, 'is the endowment by its
real head of the unit of filiation . . . the expression of the real constitution of Trobriand kinship
grouping. . . .' Hence the conclusion emerges that 'if the urigubu is regarded as a gift from an
outsider ... it appears absurd . . . but ... as the endowment of his own kindred group by its
head, it becomes natural, almost obvious'. But there is still the essential juridical point to come.
For this we have to wait till Malinowski reverts to the harvest gift in the chapter on land tenure.
There he emphasizes that though a woman is bound to go and live with her husband she
'legally remains a member of her own subclan'. This means that the urigubu can be regarded as
the 'annual return from the joint patrimony . . . due to the woman from her brother', a
consequence of the fact that 'she and her children are real owners of the soil' on which it is
raised. If this principle is grasped all the complicated details and purported exceptions fall into
place. The incidence of rights and obligations determined by the bonds of kinship and affinity is
seen to be regular almost to the point of rigidity. By this analysis it is, for instance, quite
reasonable for a woman's son to assume the uriguhu obligation if she has no brother. The
caption 'Hunger, Love and Vanity as Driving Forces in the Trobriand Harvest Gift' (1935, Vol. I,
chap. VI, sect. 2) is seen to refer to psychological elements that are only secondary to the jural
compulsion. Of course Malinowski himself did not perceive or admit this, but his analysis
leaves the issue in no doubt. Equally instructive is a reconsideration of Malinowski's favourite
topic of the gifts given by fathers to their sons. It illustrates very well how, being engrossed in
the descriptive detail, he failed to follow up the analysis in its theoretically more important
juridical aspects. What he constantly emphasizes is that these gifts of magic, material property,
and legal and political privileges are freely and affectionately made; and the implication is
always that they are outright gifts. The subject is discussed passim in Coral Gardens. The
following statement (1935, Vol. I, p. 205) is typical: 'The father, in actual fact, always tries to
give as much as he can to his own sons at the expense of those of his sister, who are his legal
heirs. His natural inclinations are seconded by customary usage which almost defies and
certainly circumvents the rigid matrilineal law, by giving the father a number of opportunities
to favour his sons and to curtail the rights of his matrilineal nephews.' The narrative goes on to
contrast the gift to a son with the nephew's obligation to 'buy' the magic from his uncle. The
interpretation follows earlier lines. Thus the gifts are described (p. 360) as *a dynamic
adjustment between the patriarchal and the matriarchal principle'; but certain qualifications
are now made more explicit. We learn more about the way in which rank tips the balance in
favour of the son as against the nephew. It is in the interest of a chief to make gifts to his son
and he can do so, whereas a commoner lacks the means and authority to do so. But an
important point emerges. The benefit of gifts is associated with the rules of marriage and
residence. A son gains by them as long as he remains in his father's village. But this cannot
outlast a generation since sons move to their matrilineal villages in adulthood. Even if the son
marries his father's sister's daughter, as chiefs desire, and stays on in his father's village in
uxorilocal marriage, his oflfspring automatically become full-titled citizens and the introduced
lineage is eliminated (p. 364). If he marries an alien his children go off to their maternal kin,
and the outcome is the same. In the first case the son is an 'intermediate heir' (p. 205)
between his father and his son, who will, of course, inherit from the father's sister's son. But
the point here is that the first recipient of the gift, by an act of fatherly gift-giving to his son in
the second generation, restores the inheritance from which he has had an illegal 'cut' to where
it rightly belongs by descent. This is the ideal case, for in the second one the gifts are in the end
a dead los3 to both the father's and the son's lineages. The concept of an 'intermediate heir' is
of great interest. It is pre- sumably metaphorical since the essence of Malinowski's whole
analysis is the antinomy between the nephew's legal rights to inherit and the son's exclusion
from these rights. However the implication is clear. It is that what every man (especially chiefs)
tries to do is to reconcile his love for his son with his legal duty to his heir. The solution, via an
arranged cross-cousin marriage, is for the magic and property held back from the lineage for
the son, to be returned to it through the heir of the next generation. This, incidentally, is a
rationalization of cross-cousin marriage that is common among tribes with matrilineal descent
systems. The crux, here, however, is the implication that the gifts given outright to a son return
to the true heirs-by-descent through the same medium of paternal gift-giving in the context of
the chances of marriage. The very action by which the laws of inheritance, are 'circumvented'
turns round on itself and redresses the balance in the next generation. It is just the kind of
paradox Malinowski loved, though it would, of course, only work, if he is right, in the relatively
unusual case of cross-cousin marriage. What eventually happens to the son's 'introduced'
lineage is irrelevant and the statement that it is 'automatically eliminated' by his marriage is, in
fact, quite beside the point. For it is only a man's daughters who can, by uxorilocal marriage in
their father's village, if that were possible, there perpetuate the lineage to which his sons
belong. This is not the place to follow out the further ramifications of Malinowski's argument.
What I want to seize on is that his picture represents a father's gifts as outright alienation
contrary to the laws of inheritance but inevitably bound to revert to the true owners after a
generation as a result of further gift giving, the motive force behind this movement being the
personal relationships of fathers, sons, uncles and nephews. He is so determined on this that
when he records how magical leader- ship was passed on at Omarakana for three successive
generations from father to son, it is almost as an afterthought that he mentions the last
holder's status as a legitimate heir to these powers in virtue of his father's cross-cousin
marriage. But what is also striking is that paternal gifts, whether of magic or of land or of
privilege, benefit the recipient mainly, if not only, while he is a resident in his father's village.
Localized or not, these benefits are bound to terminate with the son not for personal reasons,
but because the combination of the rules of descent, marriage, and residence makes it
impossible for them to be transmitted to the son's heirs and, in so far as they are locaHzed,
useless to hand on to his sons unless they are village citizens by descent. This suggests an
alternative to Malinowski's interpretation of the data and one that is, in my opinion, more
consistent with the total social structure of the Trobriands. On this view we should regard the
gifts as being in reality a sharing with his sons, by the father, of his possessions and rights on
the same principle as a child is permitted and indeed entitled to share any food offered to his
father, whatever its source. This sharing, or lending, is for the time being only, and holds only
during the father's lifetime. It arises quite normally out of the father's duty to rear his children
to adulthood; and it is not a 'circumvention' of the laws of inheritance. On the contrary, it is an
aspect of the rightful employment by a father, in his capacity of legal holder for his lifetime, of
any properties or privileges that accrue to him as a member of his lineage and as his uncle's
heir. While he is the holder he is entitled to use his inheritance as he pleases, provided that his
heir is not deprived when he in turn comes to inherit. To allow his children to use and enjoy his
inheritance during his lifetime—perhaps only during their minority —is both his pleasure and
his duty as a father. When he dies the rights of the lineage, in the person of his heir, are
immediately reasserted. The whole estate, including the portions of which the sons have had
temporary benefit, reverts to the Hneage by rights of inheritance. This explains why the so-
called gifts do not outlast a generation and fits in well with Trobriand rules of descent,
inheritance, marriage and residence. I should be willing to argue that some such interpretation
must be right on first principles; for constant, even frequent, evasion of the rules governing the
control over property and legal privilege and the trans- mission of rights between successive
generations would not be tolerated in any society. But confirmation for it is found in
Malinowski's own data, though in an unexpected quarter. In The Sexual Life of Savages (1929a,
p. 178), a contrast is drawn between the nephew's payment of pokala tribute to his uncle in
order to establish his claim to his inheritance, and the 'gratis' gifts to the son, received 'without
the sanction of tribal law'. And then comes this sentence: 'Of course he [i.e. the son] has to
return them, at least in part, after his father's death; but the use and enjoyment he has had of
the material benefits remain his, while the magic he cannot return.' It is a remark that raises
many questions. What part of the gifts must the son return? How is this enforced, as it
obviously is, by tribal law? If much magic is given away to sons how is the right to it that we
know (cf. p. 180 above) to be vested in the descent group maintained at all? Does a son's
knowledge of the magic automatically confer on him the right to utilize it freely or can he only
practise it with the consent of his father's heir? Or perhaps some kinds of magic are a form of
'liquid' resources that do not accrue to a heritable estate. These questions receive no direct
answer in the Trobriand corpus, for they are significant only in relation to the kinship system
viewed as a total system of primarily jural relationships. This Malinowski never did. Indeed he
barely touches on kinship relations outside those of parents. The relationships of grand-parents
and grand-children, for example, who appear, from the terminology, to be closely identified,
are only incidentally mentioned.

My conclusion is that it is, in fact, impossible for a Trobriand father to endow his sons with gifts
at the expense of his matrilineal kin. Indirect evidence and the sentence just quoted support
this hypothesis. The interpretation I have proposed would explain why tribal usage appears to
accept with tolerance a father's partiality to his children. It would explain the pokala as a
formal, jural institution by which the rights of the heir are affirmed, almost as a reminder to a
man that he has not got the freedom to give away what belongs to the lineage. It is consistent
with Malinowski's own statements about cross-cousin marriage, and about the relations of
affines. A chief's son first enjoys local benefits by sharing them with his father, and continues to
enjoy them for another generation in virtue of his privileged relationship with his father's heir
(his cross- cousin wife's brother) and perhaps as a sort of trustee for his son, the next heir. It
shows Malinowski's drama of father-love pitted against the 'ideals' of the law in the more
acceptable perspective of conduct for which provision is made in the law itself. The evidence I
have quoted shows that Malinowski was aware of this aspect of the problem. His failure to
follow it up, especially in Coral Gardens which is so largely concerned with questions of the
ownership and inheritance of property and of other legal rights, is attributable, I believe, to his
theoretical bias. But it is characteristic of his ethnographic insight and objectivity that the facts
are recorded in such a way as to make it possible for us to find in them the answers to our
questions. I have, in this essay, taken the liberty of criticizing some of Malinowski's ideas and
hypotheses. The criticisms I have made seem to me to be well founded when we examine
Malinowski's work from the point of view of the structural theory which is yielding such
valuable results in the present stage of kinship research. I have suggested that Malinowski's
conceptual frame was not adapted to the study of kinship systems in their own right, as a part
of the total social structure. But that it is important and extremely profitable to re-examine
Malinowski's contri- bution to our knowledge of kinship institutions is beyond question. It is
worth criticizing his ideas just because they are still vital and relevant. By contrast, the theories
of Rivers and of those who agreed with him are merely of historic interest. Malinowski's
virtuosity as an ethnographer partly explains his continuing significance. But there are other
reasons as well. For one thing, there are the rules of procedure in field research which we owe
to his example. His insistence that our primary data must come from living societies is
pertinent par excellence to kinship studies. We now regard it as indispensable for such studies
to be based on the observation of conduct and action, thought and feeUng, among people
rather than on the interrogation of selected informants. This applies particularly to the data
used in the ethnographic monograph, which has become the standard medium of contributing
to kinship theory by means of intensive field-work. Such monographs now regularly follow the
'life- history' pattern laid down by Malinowski. In addition to Firth's book (1936) there are
studies like Schapera's Married Life in an African Tribe (1939), my own The Web of Kinship
among the Tallensi (1949a), and Evans-Pritchard's Kinship and Marriage among the Nuer (195
16), which show the influence of Malinowski's ethnographic techniques. More important are a
number of postulates about the nature and sources of kinship custom and conduct which we
owe chiefly to Malinowski. No study of kinship, whether it is purely ethnographical or analytical
or comparative, can now ignore the part played by sexual institutions and values. Murdock
(1949) shows this by devoting a detailed chapter of his comprehensive comparative survey of
kinship and social structure to the 'regulation of sex'. Even if we think of sexual tendencies and
habits as belonging to the biological factors which set limits to kinship forms, rather than as
directly constitutive of them, we can no longer overlook them. And when anthropologists
describe the sexual elements in social organization they do so in terms of Malinowski's general
approach, and often of his re-interpretation of psycho-analytical hypotheses. Thus Evans-
Pritchard, writing of Nuer marriage (195 16, p. 49), says: 'Sexual activities are from their
earliest manifestations given the stamp of cultural values. They are from the first associated
with marriage, which is the final goal of the sex life of men and women.' There follow in
succession descriptions of children's sexual play, the flirtations of adolescents, the love affairs
leading to courtship and, finally, marriage. In this connection, too, it is now usual to describe
the beliefs held about the nature of procreation and to show their relationship to family and
kinship organization (cf. Schapera (1939) and Fortes (1949a)). Here we must give particular
recognition to the problem of incest and exogamy, which is a central topic in kinship studies. In
my opinion there is no evidence of any advance on the theory first sketched out with such
penetrating insight in Sex and Repression (1927a). Take LeviStrauss's disquisition on the subject
in his erudite and complex study of cross-cousin marriage (1949). The title of his first chapter,
'Nature et Culture', reminds us of the heading in Part IV of Sex and Repression ('The Transition
from Nature to Culture'); and the conclusion that it is the incest prohibition that marks the
transition (p. 13) accords with Malinowski's. His theory of incest-cum-marriage-regulation—for,
like Malinowski, he confounds the two—is not, in my judgement, tenable. 'The prohibition of
incest', he sums up (p. 596), 'is less a rule that for- bids marrying the mother, sister or daughter
than a rule which obliges the giving of the mother, sister or daughter to another. It is the rule of
the gift par excellence.' Taken literally, this conflicts directly with native statements reported by
the majority of ethnographers and with our own notions. The incest taboo is universally
thought of and stated as the prohibition of sexual relations between specified kin. This
argument, and the case made by Levi-Strauss against Malinowski's theory, in my view have no
ethnographic basis. The common opinion in anthropology and sociology (which, as I have
indicated, I do not myself share) is better shown in Murdock's statement that *. . . incest
taboos and exogamous restrictions of whatsoever sort seem clearly to be extensions of the sex
taboos between parent and child and between brother and sister' (1949, p. 284). This is pure
Malinowski. More to the point, perhaps, is the testimony of Parsons in a recent article (1954, p.
102). He writes: 'The universality of some order of incest taboo is of course directly connected
with the fact that the nuclear family is also universal to all known human societies. The
minimal criteria of the nuclear family are, I suggest, first that there should be a solidary
relationship between mother and child lasting over a period of years and transcending physical
care in its significance. Secondly, in her motherhood of this child the woman should have a
special relation- ship to a man outside her descent group who is sociologically the "father" of
the child, and that this relationship is the focus of the "legitimacy" of the child, of his
referential status in the larger kinship system'. Leaving aside the phrase italicized by Parsons, to
which no satisfactory anthropological meaning can be attached, this is a reasonable summary
of Malinowski's notion of the family almost in words he might have used. The extent to which
Malinowski's hypotheses have been incorporated in current social science is best seen in the
postulate that the nuclear family is the source and mainspring of all kinship custom and
behaviour. 'The point of departure for the analysis of kinship is the nuclear family,' says
Murdock (1949, p. 92), citing Malinowski in support. He goes on to state that the 'developing
child's . . . behaviour in these primary intra-family relationships tends to be extended or
"generalised" ', citing Evans-Pritchard in support. I have indicated why I do not accept
Malinowski's hypotheses about the nuclear family but most anthropologists would agree with
Murdock's restatement of it rather than with my criticisms. I suspect, however, that they would
regard the specific notion of the Initial Situation as expressing a truism rather than a law of
social organization. The Principle of Legitimacy implied in Parsons's remarks is another thing.
Ethnographic evidence seems to support it. The same idea, without the implications of the
assumption that it is related to the necessity for the mother and child to have a male protector,
underlies Radcliffe-Brown's distinction between the genitor, or physical father, and the pater, or
jural father. There is no doubt that the recognition of this principle has led to important
discoveries about kinship.^ It was the stroke of inspiration in interpreting the position of the
father in the Trobriand family that confirmed this principle for Mallnowski, though he
foreshadowed it in the AustraUan book. It is undeniable that his isolation of the 'split' in the
father's roles, and his description of how the critical features of the opposed roles, are dis-
tributed in the matrilineal family, gave anthropology a new perception of family organization.
Some anthropologists now interpret the ex- pression of this split in custom and belief as
deriving from the structural arrangements rather than from emotional conflict. It might
properly be argued that the two points of view are complementary not antithetical. In any
case, the root of the matter is Malinowski's analysis. In the current emphasis special attention
is fixed on the systems of rights and duties represented in kinship systems; and in analysing
these, a place of cardinal importance is given to rules of reciprocity. Malinowski's elucidation of
the part played by reciprocity in social life is one of his most notable contributions to modern
theory. The distinction between freely given gifts and services and those rendered as legally
binding obligations is essential to modern theory. As is well known, Malinowski's analysis of gift
giving among the Trobrianders was enthusiastically hailed by Marcel Mauss in the famous Essai
sur le Don (1925). He singled out the analysis of the husband's gifts to his wife in return for her
sexual services as a *tres grande decouverte'. LeviStrauss has built his theory of the exchange
basis of cross-cousin marriage on Mauss's concept of the gift and it is not unreasonable to
connect his theory, albeit indirectly, with Malinowski's 'great dis- covery'. Criticism of this
discovery does not diminish its importance in stimulating the development of theory. But
perhaps, when all is said and done, Malinowski's most valuable legacy to us is the fabric itself,
as he might have said. I have had to pull it apart for discussion, but the fabric into which all the
separate strands of thought and observation and intuition were woven by him is unique. It is
from the intimate study of Malinowski's work that we have learnt to see the inter-relatedness
and coherence of all those numerous components of social life to which he gave the name of
culture; and the eff"ect on ethnographic craftsmanship and theoretical insight is conspicuous in
the study of kinship and social organization. It is bound up with his revelation of the seamy side
of culture as an inevitable and socially meaningful concomitant of the life of a society,
reflecting its vital purposes and giving strength to it. What Malinowski revealed was that the
deepest layers of conduct, feeling and social relationship are manifested in custom and are
therefore accessible to scientific inquiry without overstepping the bounds of ethnographic
method. I do not think we have fully assimilated this discovery. Perhaps the book on The
Psychology of Kinship is still waiting to be written by somebody who will do so.